The 1988 recommendations from the Joint National Committee revolutionized the step-care approach to the treatment of hypertension: the calcium- channel blockers, a new class of agents, are now included as first-line agents. That they lower blood pressure is clear. But no randomized trial has yet evaluated their safety and efficacy in terms of the clinical endpoints of stroke and coronary heart disease (CHD). Because the results of previous randomized trials suggest that the treatment of hypertension does not reduce the incidence of CHD, we have designed a population-based, case-control study to determine whether the calcium- channel blockers reduce the incidence of myocardial infarction (MI) in patients with hypertension. Our secondary aims include the evaluation of the relative efficacy and safety of other major classes, including ACE inhibitors, beta-blockers, and alpha blockers. This revised proposal includes the results of three pilot studies that address questions raised in previous reviews. The setting is Group Health Cooperative (GHC). We will use the GHC computerized files to identify potential cases: all treated hypertensive patients, aged 30 to 79 yrs., will be eligible as cases if, according to WHO criteria, they present with an incident, fatal or non-fatal MI. A random sample of members listed in the GHC enrollment files will serve as our source of potential controls. Review of the outpatient medical records will ensure that all study subjects meet the same entrance criteria. This effort will also secure information about blood pressures, duration of hypertension, and past medical history. A telephone interview will provide information about other potential confounders, including smoking, diet, and physical activity. The GHC computerized pharmacy records, a database of all prescriptions filled by enrollees, will serve as the primary source of information about the use of calcium-channel blockers. Frequency matching will control for the potential confounding effects of age and year of presentation, and data analysis will involve logistic regression. Collecting data for 7 years will identify a number of study subjects sufficient to provide 80% power for detecting a relative risk of 0.70. The use of population-based controls will enhance the validity of our findings--results which will provide evidence about the overall risk or benefit from the use of calcium-channel blockers and other anti- hypertensive agents in patients with high blood pressure.